A Novel Adaptive Recurrent Neural Network-Based Control Strategy for Type 3 WF to Mitigate Subsynchronous Interactions

The interaction between the wind farm (WF) and the series-compensated long transmission line may cause subsynchronous interaction (SSI) phenomenon. The conventional subsynchronous damping controller (SSDC) is particularly notable for its widespread acceptance, primarily attributed to its simple implementation. However, owing to the nonlinear behavior of wind-integrated power systems under uncertain operating conditions, the effectiveness of the conventional SSDC deteriorates. Data-driven methods offer alternative solutions to this issue. This article proposes a novel adaptive control technique to suppress SSI in doubly fed induction generator (DFIG)-based wind power systems using recurrent dual hidden layer radial basis function neural network (RDHL-RBFNN). RDHL-RBFNN is used to approximate nonlinear functions, encompassing the SSI component within the DFIG. Closed-loop system stability is demonstrated using the Lyapunov method. The test bed system is built in MATLAB/Simulink to replicate the sub synchronous resonance event 2009 in the Electric Reliability Council of Texas. The time domain and controller hardware in loop simulations demonstrate the effectiveness of the proposed RDHL-RBFNN-based supplementary controller and its superior damping characteristics compared to the conventional methods to suppress SSI. To further establish the effectiveness of the proposed controller, it is tested in an interconnected WF model, and its capability to mitigate SSI is verified.